lokeshj1703 commented on a change in pull request #1604:
URL: https://github.com/apache/ozone/pull/1604#discussion_r529514509
##########
File path:
hadoop-hdds/server-scm/src/main/java/org/apache/hadoop/hdds/scm/block/DeletedBlockLogImpl.java
##########
@@ -129,9 +129,12 @@ public void incrementCount(List<Long> txIDs) throws
IOException {
DeletedBlocksTransaction block =
scmMetadataStore.getDeletedBlocksTXTable().get(txID);
if (block == null) {
- // Should we make this an error ? How can we not find the deleted
- // TXID?
- LOG.warn("Deleted TXID {} not found.", txID);
Review comment:
@xiaoyuyao I can see this warning message being printed a lot of times
during deletion. It can easily fill up SCM log during deletion. How about we
change the log level and investigate the reason in a separate jira?
SCM is processing thousand of transactions and it can resend these
transactions to a particular datanode in case ACK is delayed. On receiving the
ACK it will remove the txns from the table and while sending the transactions
which are already in queue it will try to fetch the transaction from the table.
This can lead to warning being printed thousands of times.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
